Cognitive bias and diagnostic overshadowing in trauma: The case of substance use disorder in patients with traumatic brain injury.

BACKGROUND

Diagnostic overshadowing is a type of cognitive bias where a pre-existing condition overshadows the evaluation of other diagnoses. We aimed to explore diagnostic overshadowing in trauma, using as a proxy the effects of substance and alcohol use disorder on time to surgical intervention in patients with traumatic brain injury.

METHODS

The 2017-2020 American College of Surgeons-Trauma Quality Improvement Program database was used to identify patients of at least 18 years of age with isolated traumatic brain injury who underwent craniotomy, intraventricular drain, or intracranial pressure monitoring. The primary analysis included patients with Glasgow Coma Scale of at least 13 since diagnostic overshadowing is more likely in patients with mildly altered sensorium. Patients were stratified into those with and without substance use disorder. Multivariable logistic regression examined the effects of substance and alcohol use disorder on delayed surgery (>48 h) and outcomes (eg, mortality, sepsis). Sensitivity analyses were conducted for patients with Glasgow Coma Scale below 13.

RESULTS

Of 4,005,762 trauma patients, we included 15,207, of which 4,856 had Glasgow Coma Scale of 13 or higher: 869 (18%) substance and alcohol use disorder. The mean time to procedure was 46.7 ± 108 hours in substance and alcohol use disorder compared with 32.4 ± 65 hours in nonsubstance and alcohol use disorder (P < .001). On multivariable analyses, patients with substance and alcohol use disorder were more likely to experience delayed surgery (adjusted odds ratio [aOR], 1.31; 95% CI, 1.06-1.63), mortality (aOR, 1.36; 95% CI, 1.08-1.7), postoperative sepsis (aOR, 2.44; 95% CI, 1.34-4.41), and prolonged hospital stay (aOR, 4.04; 95% CI, 2.76-5.32) compared with non-substance use disorder. Sensitivity analyses showed no difference in time to procedure or outcomes in substance use disorder and nonsubstance and alcohol use disorder patients with Glasgow Coma Scale below 13.

CONCLUSION

Diagnostic overshadowing can be measured and negatively affects patient care. Further studies are needed to explore the prevalence and effects of cognitive bias in patients with trauma.